Budget: 700 UAH Deadline: 7 days
Hello, I recently worked on a similar high-load request sender: a connection pool, asynchronous workers, batching, and rate limit control for 100-200k requests per minute. Retries with backoff and a queue for failed requests were part of the same solution.
If requests are going to an external API with limits, it's worth implementing a queue and adaptive throttle right away, as this saves from bans and data loss under peak load.
Where exactly are the requests being sent - to your own server or an external API, and does it have a limit on the number of requests per minute? And where is this supposed to run - a one-time launch or a continuous service?
I suggest we get in touch; I can also sketch out a solution diagram with a queue and workers for your 200k per minute.